Language Language English 简体中文

unilab.envs.mdp.actions.actions

Joint transmission actions for the NumPy Manager-Based runtime.

Classes

BaseAction	Apply a cold-path-resolved affine transform to raw policy actions.
BaseActionCfg	Configuration shared by entity joint actions.
JointEffortAction	Control joints via effort targets.
JointEffortActionCfg	Configuration for joint effort (torque) control.
JointPositionAction	Convert policy actions into entity joint-position targets.
JointPositionActionCfg	Configuration for joint-position control.
JointVelocityAction	Control joints via velocity targets.
JointVelocityActionCfg	Configuration for joint velocity control.
RelativeJointPositionAction	Control joints via position targets relative to current positions.
RelativeJointPositionActionCfg	Joint position targets relative to the current measured position.

class unilab.envs.mdp.actions.actions.JointEffortAction

Bases: BaseAction

Control joints via effort targets.

Parameters:

• cfg (BaseActionCfg)

• env (ManagerBasedRlEnv)

apply_actions()

Return type:

None

class unilab.envs.mdp.actions.actions.JointEffortActionCfg

Bases: BaseActionCfg

Configuration for joint effort (torque) control.

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

build(env)

Build the action term from this config.

Parameters:

env (ManagerBasedRlEnv)

Return type:

JointEffortAction

__init__(*, entity_name, clip=None, actuator_names, scale=1.0, offset=0.0, preserve_order=False)

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

class unilab.envs.mdp.actions.actions.JointPositionAction

Bases: BaseAction

Convert policy actions into entity joint-position targets.

Parameters:

• cfg (JointPositionActionCfg)

• env (ManagerBasedRlEnv)

cfg: JointPositionActionCfg

__init__(cfg, env)

Parameters:

• cfg (JointPositionActionCfg)

• env (ManagerBasedRlEnv)

apply_actions()

Return type:

None

class unilab.envs.mdp.actions.actions.JointPositionActionCfg

Bases: BaseActionCfg

Configuration for joint-position control.

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

• use_default_offset (bool)

use_default_offset: bool = True

build(env)

Build the action term from this config.

Parameters:

env (ManagerBasedRlEnv)

Return type:

JointPositionAction

__init__(*, entity_name, clip=None, actuator_names, scale=1.0, offset=0.0, preserve_order=False, use_default_offset=True)

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

• use_default_offset (bool)

actuator_names: tuple[str, ...] | list[str]

entity_name: str

Name of the entity in the scene that this action term controls.

class unilab.envs.mdp.actions.actions.JointVelocityAction

Bases: BaseAction

Control joints via velocity targets.

Parameters:

• cfg (JointVelocityActionCfg)

• env (ManagerBasedRlEnv)

__init__(cfg, env)

Parameters:

• cfg (JointVelocityActionCfg)

• env (ManagerBasedRlEnv)

apply_actions()

Return type:

None

class unilab.envs.mdp.actions.actions.JointVelocityActionCfg

Bases: BaseActionCfg

Configuration for joint velocity control.

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

• use_default_offset (bool)

use_default_offset: bool = True

build(env)

Build the action term from this config.

Parameters:

env (ManagerBasedRlEnv)

Return type:

JointVelocityAction

__init__(*, entity_name, clip=None, actuator_names, scale=1.0, offset=0.0, preserve_order=False, use_default_offset=True)

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

• use_default_offset (bool)

actuator_names: tuple[str, ...] | list[str]

entity_name: str

Name of the entity in the scene that this action term controls.

class unilab.envs.mdp.actions.actions.RelativeJointPositionAction

Bases: BaseAction

Control joints via position targets relative to current positions.

Parameters:

• cfg (BaseActionCfg)

• env (ManagerBasedRlEnv)

apply_actions()

Return type:

None

class unilab.envs.mdp.actions.actions.RelativeJointPositionActionCfg

Bases: BaseActionCfg

Joint position targets relative to the current measured position.

target = current_joint_pos + action * scale. A fixed offset has no useful meaning for this transmission and is rejected during construction.

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)

build(env)

Build the action term from this config.

Parameters:

env (ManagerBasedRlEnv)

Return type:

RelativeJointPositionAction

__init__(*, entity_name, clip=None, actuator_names, scale=1.0, offset=0.0, preserve_order=False)

Parameters:

• entity_name (str)

• clip (dict[str, tuple] | None)

• actuator_names (tuple[str, ...] | list[str])

• scale (float | dict[str, float])

• offset (float | dict[str, float])

• preserve_order (bool)